Build multi-tenant RAG with Neon's database-per-user model — no nosy neighbors, max isolation, minimal costs
Docs/Neon Postgres guides/Functions/JSON functions/jsonb_extract_path_text

Postgres jsonb_extract_path_text() Function

Extracts a JSON sub-object at the specified path as text

The jsonb_extract_path_text function is designed to simplify extracting text from JSONB data in Postgres. This function is similar to jsonb_extract_path — it also produces the value at the specified path from a JSONB object but casts it to plain text before returning. This makes it more straightforward for text manipulation and comparison operations.

Try it on Neon!

Neon is Serverless Postgres built for the cloud. Explore Postgres features and functions in our user-friendly SQL editor. Sign up for a free account to get started.

Sign Up

Function signature

jsonb_extract_path_text(from_json JSONB, VARIADIC path_elems text[]) -> TEXT

The function accepts a JSONB object and a variadic list of elements that specify the path to the desired value.

Example usage

Let's consider a users table with a JSONB column named profile containing various user details.

Here's how we can create the table and insert some sample data:

CREATE TABLE users (
    id INT,
    profile JSONB
);

INSERT INTO users (id, profile)
VALUES
    (1, '{"name": "Alice", "contact": {"email": "alice@example.com", "phone": "1234567890"}, "hobbies": ["reading", "cycling", "hiking"]}'),
    (2, '{"name": "Bob", "contact": {"email": "bob@example.com", "phone": "0987654321"}, "hobbies": ["gaming", "cooking"]}');

To extract and view the email addresses of all users, we can run the following query:

SELECT id, jsonb_extract_path_text(profile, 'contact', 'email') as email
FROM users;

This query returns the following:

| id | email              |
|----|--------------------|
| 1  | alice@example.com  |
| 2  | bob@example.com    |

Advanced examples

Use output of jsonb_extract_path_text in a JOIN clause

Let's say we have another table, hobbies, that includes additional information such as difficulty level and the average cost to practice each hobby.

We can create the hobbies table with some sample data with the following statements:

CREATE TABLE hobbies (
   hobby_id SERIAL PRIMARY KEY,
   hobby_name VARCHAR(255),
   difficulty_level VARCHAR(50),
   average_cost VARCHAR(50)
);

INSERT INTO hobbies (hobby_name, difficulty_level, average_cost)
VALUES
    ('Reading', 'Easy', 'Low'),
    ('Cycling', 'Moderate', 'Medium'),
    ('Gaming', 'Variable', 'High'),
    ('Cooking', 'Variable', 'Low');

The users table we created previously has a JSONB column named profile that contains information about each user's preferred hobbies. A fun exercise could be to find if a user has any hobbies that are easy to get started with. Then we can recommend they engage with it more often.

To fetch this list, we can run the query below.

SELECT
  jsonb_extract_path_text(u.profile, 'name') as user_name,
  h.hobby_name
FROM users u
JOIN hobbies h
ON jsonb_extract_path_text(u.profile, 'hobbies') LIKE '%' || lower(h.hobby_name) || '%'
WHERE h.difficulty_level = 'Easy';

We use jsonb_extract_path_text to extract the list of hobbies for each user, and then check if the name of an easy hobby is present in the list.

This query returns the following:

| user_name | hobby_name |
|-----------|------------|
| Alice     | Reading    |

Extract values from JSON array with jsonb_extract_path_text

jsonb_extract_path_text can also be used to extract values from JSONB arrays.

For instance, to extract the first and second hobbies for everyone, we can run the following query:

SELECT
    jsonb_extract_path_text(profile, 'name') as name,
    jsonb_extract_path_text(profile, 'hobbies', '0') as first_hobby,
    jsonb_extract_path_text(profile, 'hobbies', '1') as second_hobby
FROM users;

This query returns the following:

| name  | first_hobby | second_hobby |
|-------|-------------|--------------|
| Alice | reading     | cycling      |
| Bob   | gaming      | cooking      |

Additional considerations

Performance and indexing

Performance considerations for jsonb_extract_path_text are similar to those for json_extract_path. It is efficient for extracting data but can be impacted by large JSONB objects or complex queries. Indexing the JSONB column can improve performance in some cases.

Alternative functions

  • jsonb_extract_path - This is a similar function that can extract data from a JSONB object at the specified path. The difference is that it returns a JSONB object, while jsonb_extract_path_text always returns text. The right function to use depends on what you want to use the output data for.
  • json_extract_path_text - This is a similar function that can extract data from a JSON object, (instead of JSONB) at the specified path.

Resources

Need help?

Join our Discord Server to ask questions or see what others are doing with Neon. Users on paid plans can open a support ticket from the console. For more details, see Getting Support.

Last updated on

Was this page helpful?